Project Summary Alzheimer's disease (AD) is the most common form of dementia and is characterized by progressive cognitive impairment and neurodegeneration. Despite decades of clinical, neuropathological and neurobiological research the molecular substrates and causal biological substrates of AD remain elusive. Most studies have focused on vulnerable brain regions defined by neuroimaging or neuropathological findings. However, our recent unbiased study of transcriptomic networks in 19 brain regions identified the parahippocampal gyrus as the brain region with the greatest transcriptomic changes associated with disease development and progression irrespective of whether disease development and progression were defined using functional measures of cognition or the canonical neuropathological lesions of AD (Wang MH et al, Genome Medicine 8:104). These observations have been reinforced further by our larger scale multi-Omics data from over 200 donors and 4 brain regions, for which we have generated whole exome, whole genome and RNA sequencing data through the current AMP-AD Consortium. These studies, and most others published to date, have not examined transcriptomic, epigenetic and proteomic changes in the same tissues and donors and by utilizing whole tissue homogenates have been unable to identify vulnerabilities holistically and with cell-type-specific fidelity. Here we propose to generate additional, matched large-scale proteomic and epigenetic data as well as cell type specific transcriptomic and epigenomic data from the parahippocampal gyrus and develop novel network inference and analysis approaches to integrate all these multi-Omics data as well as cognitive, pathological and physiological data to construct high-resolution, multiscale molecular networks in the parahippocampal gyrus in AD. To overcome some of the drawbacks of postmortem studies and to gain insight into causal mechanisms and networks we will experimentally validate networks identified in postmortem tissue by perturbing a number of key drivers in AD transgenic mouse primary brain cells and human iPSC derived brain cell cultures to determine the mechanisms underlying vulnerability of the parahippocampal gyrus. We expect that our data-driven and hypothesis-free multiscale network modeling of parahippocampal vulnerability in AD will have a large impact on the AD field and lead towards a more comprehensive and precise understanding of AD pathogenesis. More importantly, the proposed research will pave a path towards drug discovery for AD targeted at specific vulnerabilities in specific brain cell types.